.notna() vs .notnull()

.notna() vs .notnull() in pandas

Just like before, these two are also identical — both check for non-missing values (i.e., values that are not NaN, None, or NaT).

---

1. Purpose

They return True where the data exists (not missing).

Let’s use the same example:

import pandas as pd
import numpy as np

data = {
    'name': ['Dipak', 'Sita', 'Ram', None],
    'age': [25, np.nan, 30, 28]
}

df = pd.DataFrame(data)
print(df)

Output:

    name   age
0  Dipak  25.0
1   Sita   NaN
2    Ram  30.0
3   None  28.0

---

2. Using notna()

print(df.notna())

Output:

    name    age
0   True   True
1   True  False
2   True   True
3  False   True

---

3. Using notnull()

print(df.notnull())

Output (identical to notna()):

    name    age
0   True   True
1   True  False
2   True   True
3  False   True

---

Comparison Table

Function	Meaning	Returns True When	Same As	Use For
isna()	is missing	Value is NaN / None / NaT	isnull()	Finding missing values
notna()	is not missing	Value exists	notnull()	Keeping valid values

---

Quick Summary

Missing?	Function	Returns
Yes	isna() / isnull()	True
No	notna() / notnull()	True

---

Practical Examples

Select only rows with missing data

df[df.isna().any(axis=1)]

Select only rows with valid (non-missing) data

df[df.notna().all(axis=1)]

Count missing values

df.isna().sum()

---

Conclusion

• isna() ≡ isnull() → detect missing values
• notna() ≡ notnull() → detect non-missing values
• Prefer isna() and notna() (modern naming, consistent with NumPy)

---